Professional Documents
Culture Documents
2
ISSN 1539-0578 pp. 225230
Reviewed work:
Reviewed by
Zahir Mumin
University at Albany, SUNY
United States
http://www.cambridge.org/
Khalifa and Weir present Examining Reading in order to address controversial issues concerning
the structure and content of Cambridge English for Speakers of Other Languages (ESOL) exams.
In this 29th volume of the Studies in Language Testing (SiLT) series, the authors focus on the
dynamic relationships between students cognitive skills, the tests scoring criteria, and its
assessment tasks that exhibit a wide variety of different contexts. Khalifa and Weir
fundamentally contend that test developers must clearly explicate and attentively examine how
reading comprehension expectations align with students learning needs in order to create
reliable scoring assessments.
The book is organized into eight chapters that emphasize the importance of exam validity:
Chapter 1 Introduction; Chapter 2 Test-taker characteristics; Chapter 3 Cognitive validity;
Chapter 4 Context validity; Chapter 5 Scoring validity; Chapter 6 Consequential validity;
Chapter 7 Criterion-related validity; and Chapter 8 Conclusions and recommendations. The
authors also provide six appendices (Appendix A-F), which include reading comprehension tasks
that substantiate their main argument in favor of meticulous learner outcome expectations.
Chapter 1 analyzes Weirs (2005) theoretical framework for conceptualizing reading test
validity (p. 5, Figure 1.1) to demonstrate how this framework effectively employs a socio-
http://nflrc.hawaii.edu/rfl
Mumin: Review of the book Examining Reading 226
cognitive approach by associating validity components with students mental processes as they
complete reading comprehension tasks. In this chapter, Khalifa and Weir contextualize the
content of Chapters 27, which further develop and apply Weirs theoretical framework to
Cambridge ESOL exam constructs. Chapter 8 summarizes Chapters 17 in chronological order.
In Chapter 1, Khalifa and Weir foreground the content of Chapters 27 by defining the six
validity components named in the chapter titles of the book. Test-taker characteristics consider
how students learning needs and sociocultural background affect their ability to complete
different tasks. Context validity encompasses the sociocultural and sociolinguistic contexts in
which tasks are performed. Cognitive validity examines the extent to which reading tasks are
relevant to authentic language use. Scoring validity determines if the evaluation procedures of
assessments are quantitatively consistent and reliable. Consequential validity explores how bias,
washback, and social values affect testing procedures. Criterion-related validity analyzes the
integration of test scores and external criteria to infer reliability of assessment procedures. The
authors assert that the intense interaction amongst these components significantly impacts
students reading proficiency levels.
Chapter 2 outlines OSullivans (2000) categories of test-taker characteristics (p. 18, Table
2.1) to investigate the physical/physiological, psychological, and experiential factors that may
influence ESOL students performance on reading proficiency exams. Khalifa and Weir contend
that individual personal characteristics of test takers such as learning disabilities, motivation, and
exam preparedness greatly impact their performance on reading tasks. The authors initiate the
defense of this argument through analysis of international and the U.S. test accommodation
policies which permit various types of exam modifications for students with learning disabilities
(physical/physiological factors), such as extended time, Braille, scribers, readers, and different
environmental settings without distractions. Khalifa and Weir focus on extended time and Braille
as two of the most important accommodations that relate to students ability to complete reading
tasks.
Khalifa and Weir note that Sireci, Scarpati, and Li (2005) spark insightful controversy in the
field of second language reading with their review of empirical studies which show that extended
time enhances the academic disposition of students with disabilities giving an advantage over
those without disabilities. However, Khalifa and Weir also point to Prings (1994) findings
concerning Braille readers which demonstrate that they read at a much slower pace than print
readers and therefore, should be permitted extended time. The authors provide this example of
extended time use to allay the controversy regarding the possible unfair advantage for students
with learning disabilities. Khalifa and Weir also address issues dealing with fairness in their
examination of psychological (personality) and experiential (educational background) factors.
They argue that Cambridge ESOL test development procedures effectively gather personal and
demographic information from exam candidates using Candidate Information Sheets (CIS) and
Differential Item Functioning (DIF) in an attempt to prevent bias and ameliorate the content of
reading activities. They supplement this argument with further claims that the pretest conducted
by Cambridge ESOL administrators focuses on eliciting information from students about their
previous knowledge of exam formats and popular topics of personal interest. This information, in
turn, makes exams contextually more relevant to students personal lives and prior exam
Chapter 3 investigates the importance of cognitive validity in ESOL exams. Khalifa and Weir
initiate the chapter by defining and comparing exam validation approaches such as the factorial
approach, the reading subskills approach, and the cognitive processing approach. Factorial
approaches focus on quantitatively describing specific abilities that are generally necessary for
reading comprehension success: recognizing key words, understanding the main ideas of
paragraphs, and identifying similar patterns of sentence structure. Reading subskills approaches
examine the effectiveness of different skills (automatized abilities/subconscious acquisition) and
strategies (learned abilities/conscious learning) used to comprehend texts. Khalifa and Weir
demonstrate the impact of cognitive processing approaches on exam development by creating a
functional reading model (Figure 3.1, p. 43), describing its components (types of reading and
cognitive processes), and applying these components to exam constructs of Cambridge ESOL
Main Suite Reading papers from the Key English Test (KET), Preliminary English Test (PET),
First Certificate in English (FCE), Certificate in Advanced English (CAE), and Certificate of
Proficiency in English (CPE).
The authors first analyze types of reading and confirm Ashtons (2003, p. 128) findings which
show that the gapped-text tasks on the CAE and CPE exams oblige students to employ careful
global reading skills (profound higher-order thinking skills) in order to successfully comprehend
texts. In contrast, Khalifa and Weirs examination of cognitive processes shows that the majority
of the content of the Cambridge ESOL Main Suite Reading papers tasks does not require
students to use their most complex cognitive processing skills (creating a text level structure
[and] creating an organized representation of several texts p. 70) to integrate the meaning of
sentences in different paragraphs and/or explicate the intertextual meaning throughout multiple
texts. This chapter could be enhanced with linguistic analyses of cognitive processes in order to
determine, for example, if the development of ESOL students comprehension skills of inference
are primarily due to syntactic parsing, lexical access, or word recognition.
Chapter 4 explores polemical issues related to the development of contextually valid ESOL
exams. Khalifa and Weir argue that tasks must exhibit real-life contexts associated with students
personal lives in order for exam administrators to reliably determine the extent to which students
acquire certain levels of English reading proficiency. The authors substantiate this argument by
applying the fifteen components of Weirs (2005) context validity model (p. 82, Figure 4.1) to
different exam formats (e.g., multiple choice and true/false) and specific content (e.g., grammar
and lexicon) included in the Cambridge Main Suite Reading papers exams. Of these fifteen
components, Khalifa and Weir highlight the significance of the following two components:
order items (p. 82) from the task setting category and content knowledge (p. 82) from the
linguistic demands category. For order items, the authors contend that the order in which
students are required to provide answers to reading comprehension questions should match the
order in which the relevant reading material is presented to them. Hughes (1989) and Weir
(1993) support this argument by concluding that students normally process reading material in a
chronological order. Therefore, when these students are exposed to exercises using a random
order of responses, this order may impede comprehension and degrade the reliability of test
performance results. Khalifa and Weir further solidify the aforementioned argument by
illuminating the inconsistency between the chronological order (careful reading activities) and
random order (expeditious reading activities) of responses required on Cambridge ESOL exams.
The authors conclude the chapter by arguing in favor of matching general test topics (content
knowledge) of the Main Suite Reading papers to students background knowledge in order to
stimulate their reading comprehension and enhance the authenticity of context validity.
Chapter 5 focuses on the employment of appropriate statistical analyses of exam tasks, results,
and scoring procedures, which help determine the reliability of Cambridge ESOL reading
comprehension exams. Khalifa and Weir postulate that scoring validity is one of the most
imperative components because the lack of scoring validity constitutes grave deficiencies in
cognitive validity and context validity. The authors sustain this postulate by applying the six
components of Weirs (2005) scoring validity model (p. 144, Figure 5.1) to Cambridge ESOL
scoring practices. The six components are item difficulty, item discrimination, internal
consistency, error of measurement, marker reliability, and grading and awarding.
Item difficulty and item discrimination measure the relationship between task performance, task
facility, and number of examinees. Internal consistency determines the extent to which students
achieve similar scores when identical skills are assessed. Error of measurement assures that
students with scores close to the borderline of pass/fail are not adversely affected by human
errors.
Marker reliability develops detailed procedures for maintaining scoring consistency in manually
graded exams. Grading and awarding establishes grading norms for quantifying cut-off
passing/failing grades and develops criteria for written results notifications sent to students.
Khalifa and Weir demonstrate, through their analyses of various statistical procedures such as
Rasch-based statistics, the Cronbach Alpha method, and the Standard Error of Measurement, that
Cambridge ESOL exams possess high levels of reliability. The authors summary of Chapter 5s
in-depth analysis of scoring validity clashes with Chapter 6s analysis of consequential validity
because the former is an internal validation process whereas the latter is an external validation
process.
Chapter 6 delves into prior and current research concerning the influence of consequential
validity on test development procedures. Khalifa and Weir claim that impact, the effect of tests
on society; washback, the effect of tests on teaching and learning; and potential test bias are
three key factors of exam validation which represent a tug-of-war relationship between
stakeholders who have major interests in exam criteria and formats and teachers and learners
who are often concerned with all exam conditions. The authors successfully defend this claim
through analysis of Taylors (2000) Stakeholders in the Test Community model (p. 177, Figure
6.3) within the context of Cambridge ESOL test development procedures. Khalifa and Weir also
review two recent studies, one on the washback effect of CPE textbooks and the other on the
impact of International English Language Testing System (IELTS) preparation, to show that test
administrators are attempting to create exams that are more amenable to test takers real-life
reading comprehension situations. Although these attempts reflect major improvements in the
creation of reading comprehension tasks, the authors do not propose a clear resolution for the
aforementioned tug-of-war which could be resolved through direct communication and feedback
exchanges between test takers and stakeholders. I argue that the integration of DIF analysis and
test taker/stakeholder interaction mollifies and, in some cases, eliminates test bias, negative
impact, and negative washback. The need for this integration is clearly demonstrated through
prior research such as Geranpayeh and Kunnans (2007) study which examines test-taker
characteristics using DIF analysis without addressing stakerholders superordinate logistical
exam power.
Khalifa and Weir also elucidate the effectiveness of manual alignment procedures
(familiarization, specification, standardization, and empirical validation) that Cambridge ESOL
test administrators use during their test development processes to link Cambridge ESOL
proficiency standards to those of the CEFR. However, at the same time, the authors avouch the
need for further research that empirically validates the tendency to compare English proficiency
levels of other exams to those of the CEFR. Khalifa and Weir close off the chapter with Taylors
(2004) opposing arguments addressing this unresolved CEFR alignment issue by identifying
benefits concerning the facilitated interpretation of English proficiency guidelines and risks
dealing with the oversimplification of these guidelines.
Chapter 8 summarizes the previous seven chapters in chronological order, which metaphorically
supports Khalifa and Weirs arguments dealing with Chapter 4s order items concept. The
authors maintain that the reliability of construct validitythe combination of cognitive validity,
context validity, and scoring validitymust be empirically examined and the differences
between English proficiency levels must be clearly operationalized in order to provide solid
evidence that accounts for test takers real-life reading comprehension experiences. With regard
to consequential and criterion-related validity, Khalifa and Weir reinforce the importance of
constantly cross-evaluating exam content and procedures so as to develop and/or maintain high
levels of reliability in Cambridge ESOL exams. Offering suggestions for empirical research
studies at the end of this chapter would illuminate the need for quantitative and qualitative
analyses of ESOL test takers, teachers, and stakeholders as research participants.
This book bedazzles educators, exam administrators, and language acquisition professionals by
applying a socio-cognitive theoretical framework of exam validity components to Cambridge
ESOL exams with the intention of exhibiting reliability levels of different exam constructs.
Within the context of the book, however, there is not a clear definition for test takers, (who are
not stockholders) and stockholders, (who are not test takers). It is clear that the term test takers
refers to mid/low-level stockholders (students) and the term stockholders constitutes high-level
References
Ashton, M. (2003). The change process at the paper level. Paper 1 Reading. In. C. J. Weir & M.
Milanovic (Eds.), Continuity and innovation: Revising the Cambridge proficiency in
English examination 1913-2002, Studies in language testing 15 (pp. 121164).
Cambridge: UCLES/Cambridge University Press.
Geranpayeh, A., & Kunnan, A. J. (2007). Differential item functioning in terms of age in the
Certificate in Advanced English examination. Language Assessment Quarterly, 4(2),
190222.
Hughes, A. (1989). Testing for language teachers. Cambridge: Cambridge University Press.
OSullivan, B. (2000). Towards a model of performance in oral language testing. Unpublished
PhD thesis. Berkshire, ENG: University of Reading.
Pring, L. (1994). Touch and go: Learning to read braille. Reading Research Quarterly, 29(1),
6774.
Sireci, S. G., Scarpati, S., & Li, S. (2005). Test accommodations for students with disabilities:
An analysis of the interaction hypothesis. Review of Educational Research, 75(4), 457
490.
Taylor, L. (2000). Stakeholders in language testing. Research Notes, 2, 24.
Taylor, L. (2004). Issues of test comparability. Research Notes, 15, 25.
Weir, C. J. (1993). Understanding and developing language tests. New York, NY: Prentice Hall.
Weir, C. J. (2005). Language testing and validation: An evidenced-based approach.
Basingstoke, ENG: Palgrave Macmillan.
Zahir Mumin teaches Spanish courses at the University at Albany, SUNY, USA and conducts
research in the field of linguistics. His primary research interests include L1 and L2 reading
comprehension, sociolinguistics, phonology, morphosyntax, language contact, language change,
L1 and L2 acquisition, semantics, pragmatics, multilingualism, and dialect variation. Email:
zm227418@albany.edu